pith. machine review for the scientific record. sign in

arxiv: 2503.07708 · v3 · submitted 2025-03-10 · ❄️ cond-mat.str-el · cond-mat.mes-hall· hep-lat· hep-th· math-ph· math.MP

Recognition: 4 theorem links

· Lean Theorem

Exact Chiral Symmetries of 3+1D Hamiltonian Lattice Fermions

Lei Gioia , Ryan Thorngren
Authors on Pith no claims yet

Pith reviewed 2026-05-06 21:26 UTC · model claude-opus-4-7

classification ❄️ cond-mat.str-el cond-mat.mes-hallhep-lathep-thmath-phmath.MP
keywords <parameter name="0">lattice chiral symmetry
0
0 comments X

The pith

A single Weyl fermion can live on a 3+1D lattice with an exact chiral symmetry, provided that symmetry is allowed to spread across neighboring sites rather than act site-by-site.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The longstanding obstruction to putting a single Weyl fermion on a lattice — Nielsen–Ninomiya doubling — is usually stated for symmetries that act site-by-site. This paper exhibits explicit finite-range Hamiltonian models on a 3+1D cubic lattice in which the chiral symmetry itself is allowed to span a few neighboring sites, and shows that this is enough to protect a single Weyl node (and, in a companion model, a single Weyl doublet with Witten's SU(2) anomaly). The constructions are written in the BdG formalism, where the chiral charge becomes a single-particle operator S(k) that reduces to the ordinary U(1) charge at the protected node and vanishes where unwanted nodes are removed. A matching no-go theorem shows that the generator's spectrum must be non-quantized (or the algebra non-closing), so the trade-off between on-siteness and quantization is sharp. The doublet model is recognized as the standard magnetic Weyl semimetal, but reinterpreted as carrying two exact U(1) symmetries whose UV closure is the Onsager algebra, providing translation-independent protection.

Core claim

The authors construct ultralocal Hamiltonian lattice models in 3+1D whose low-energy spectrum is a single Weyl fermion (or a single Weyl doublet) protected by an exact chiral symmetry, by allowing the symmetry generator itself to be not-on-site. For the single Weyl case the chiral charge density is a nearest-neighbor bilinear whose generator has continuous (non-quantized) spectrum — a Hamiltonian analog of the Ginsparg-Wilson symmetry. For the doublet case, two quantized U(1) generators (one on-site, one not) fail to close into su(2) and instead generate the infinite-dimensional Onsager algebra in the UV, while emanating the full SU(2) flavor symmetry with Witten's global anomaly in the IR.

What carries the argument

A BdG-formalism single-particle generator S_chiral(k) that commutes with the Bloch Hamiltonian, acts as the ordinary U(1) charge τ_z at the surviving Weyl node, and vanishes at the momenta where unwanted nodes are gapped — e.g. S_chiral(k) = ½[(1+cos k_z)τ_z + sin k_z τ_x] for the single-Weyl model, which in real space is a sum of decoupled critical Kitaev (Majorana) wires along ẑ. This wire structure makes the generator finite-range but non-compact (continuous spectrum), which is exactly what is needed to dodge the Fidkowski–Xu / Nielsen–Ninomiya-style no-gos. For the doublet, splitting this generator into two quantized pieces produces an Onsager algebra whose low-energy projection becomes

If this is right

  • <parameter name="0">A continuous chiral symmetry of a single Weyl fermion can be realized exactly on a 3+1D lattice
  • provided its charge density is allowed to be a finite-range but non-on-site bilinear with non-quantized spectrum.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • <parameter name="0">Editor's read: the construction's reliance on critical Kitaev wires hidden inside the chiral charge suggests these emanant symmetries always come with a gapless 1D structure embedded in the generator — a possible obstruction to gauging that may be sharper than the 't Hooft anomaly itself.

Load-bearing premise

The argument leans on the claim that letting the symmetry generator be finite-range-but-not-on-site (and allowing its spectrum to be continuous, or its algebra to fail to close) is a genuinely milder concession than full doubling — and that the resulting symmetry still deserves to be called "exact" rather than emergent.

What would settle it

Show that for any such ultralocal Hamiltonian with these emanant anomalous symmetries, a symmetric local perturbation can open a gap at the protected node — or, conversely, exhibit a symmetric mass term commuting with the constructed S_chiral(k) (e.g. the operator in Eq. 7) that gaps the Weyl point at k=0. Either would falsify the protection claim.

read the original abstract

We construct Hamiltonian models on a 3+1d cubic lattice for a single Weyl fermion and for a single Weyl doublet protected by exact (as opposed to emergent) chiral symmetries. In the former, we find a not-on-site, non-compact chiral symmetry which can be viewed as a Hamiltonian analog of the Ginsparg-Wilson symmetry in Euclidean lattice models of Weyl fermions. In the latter, we combine an on-site $U(1)$ symmetry with a not-on-site $U(1)$ symmetry, which together generate the $SU(2)$ flavor symmetry of the doublet at low energies, while in the UV they generate an algebra known in integrability as the Onsager algebra. This latter model is in fact the celebrated magnetic Weyl semimetal which is known to have a chiral anomaly from the action of $U(1)$ and crystalline translation, that gives rise to an anomalous Hall response - however reinterpreted in our language, it has two exact $U(1)$ symmetries that gives rise to the global $SU(2)$ anomaly which protects the gaplessness even when crystalline translations are broken. We also construct an exact symmetry-protected single Dirac cone in 2+1d with the $U(1) \rtimes T$ parity anomaly. Our constructions evade both old and recently-proven no-go theorems by using not-on-siteness in a crucial way, showing our results are sharp.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

4 major / 8 minor

Summary. The authors construct ultralocal 3+1D Hamiltonian lattice models with (a) a single Weyl fermion protected by a non-on-site, non-compact (continuous-spectrum) chiral symmetry, the Hamiltonian analog of Ginsparg-Wilson, and (b) a Weyl doublet protected by two exact U(1) symmetries — one on-site, one not-on-site — whose generators close not into su(2) but into the Onsager algebra, while emanating an SU(2) flavor symmetry with Witten's global anomaly in the IR. They also give a 2+1D analog with U(1)⋊T parity anomaly and an almost-local generator, and prove a Fidkowski–Xu-style no-go result showing that quantized, exponentially-local single-particle U(1) generators cannot carry such anomalies, so the non-quantized / non-commuting features are necessary."

Significance. If correct, the paper provides the first ultralocal Hamiltonian lattice models realizing (i) a single 3+1D Weyl fermion with a finite-range non-on-site chiral symmetry and (ii) a 3+1D Weyl doublet whose SU(2) flavor symmetry emanates from two exact lattice U(1)s generating the Onsager algebra, plus an analogous 2+1D Dirac cone with the U(1)⋊T parity anomaly. These are clean, explicit constructions that sharpen the Fidkowski–Xu and Nielsen–Ninomiya no-go theorems by exhibiting the precise loophole (non-quantized or non-commuting generators) and reinterpret the well-known magnetic Weyl semimetal in a new symmetry framework. The accompanying no-go theorem in §III, while elementary, is a useful general statement. The constructions are simple enough to be checked by inspection in the BdG formalism, which is a strength.

major comments (4)
  1. [§IV, Eq. (14)] The protection argument against translation-breaking perturbations is given in a single sentence: any Q_0,Q_1-symmetric mass m_j(k) η^z σ^j 'always commute[s] with at least one term in the original Hamiltonian, so the result is a shift'. This is the load-bearing claim that the two exact U(1)s (one on-site, one not-on-site) by themselves protect the doublet once translations are broken. The claim is plausible but not actually checked: in the folded basis e_k coupling ±K, η^z mixes the two valleys, and one must verify that no linear combination of the allowed m_j(k) η^z σ^j terms can simultaneously anticommute with the kinetic σ-structures at both valleys. Please supply either an explicit case analysis (e.g., enumerate all η,τ,σ Dirac matrices commuting with both generators in (14) and check anticommutation against k_x σ_x + k_y σ_z ± sin K k_z σ_y around each node) or a numerical demonstr
  2. [§IV / §V] The identification of the UV Onsager algebra with the IR SU(2) Witten anomaly is asserted by analogy with the 1+1D U(1)_V×U(1)_A case ([27,28]) but not derived for the 3+1D doublet. Each U(1) is individually anomaly-free; the Z_2 anomaly is supposed to arise from the non-commutation. It would strengthen the paper to demonstrate this concretely — for instance, by exhibiting an obstruction to gapping under (Q_0,Q_1) that matches the mod-2 nature of the SU(2) anomaly (e.g., a 2-doublet version that *can* be gapped symmetrically), or by computing a finite-volume invariant detecting Witten's anomaly from the lattice generators. Without this, the language 'gives rise to the global SU(2) anomaly' in the abstract is stronger than what is shown.
  3. [§III] The no-go theorem assumes (a) translation invariance, (b) free-fermion / quadratic single-particle generator S_chiral, and (c) exponential locality of S_chiral. These assumptions should be stated as a formal proposition with hypotheses, since the 'theorem' is invoked to argue that the constructions in §II and Appendices C,D are 'as good as possible'. In particular, the step 'h(k) = S_chiral(k) + P(k) … is at least exponentially localized' relies on a uniform spectral gap of magnitude 1 between the zero band and the rest; the appendix E argument does not directly cover the case where S_chiral is gapless (continuous spectrum), which is exactly the regime of the constructed models. Please clarify whether the theorem as stated assumes a gap above the zero eigenspace, and if so reconcile with the role this theorem plays in the discussion.
  4. [§II vs §III] S_chiral(k) in Eq. (7) has continuous spectrum [0,1] (it is a bundle of critical Majorana wires), so it is not strictly a U(1) generator with quantized charges — yet it is repeatedly called a 'chiral symmetry' generating an R action. The relation to the SU(2)-doublet decomposition Q_0,Q_1 in Eq. (10), where the spectrum *is* quantized but the algebra is non-abelian (Onsager), is the conceptually interesting point of the paper and deserves to be stated more sharply. As written, a reader can be confused about whether the §II model has 'a U(1) chiral symmetry' (it does not, in the standard sense) or only an R symmetry that is the diagonal of two non-commuting U(1)s.
minor comments (8)
  1. [Eq. (3) and below] The statement 'It hosts two Weyl fermions at low energy when |Δ|>1' is correct for the standard magnetic Weyl semimetal, but the parameter window for the gapped single-Weyl model is later given as |Δ|>2 (after Eq. (6)) without derivation. A short remark on how the τ_x σ_y term shifts the gap-closing condition would help.
  2. [Eq. (9)] The real-space expression for Q̂_chiral has a factor 1/2 and an h.c., but the hermiticity structure of the c†c† term is ambiguous — please indicate whether the h.c. acts on the whole bracket or only on the pairing term, ideally by writing the operator with explicit summation over r and explicit conjugates.
  3. [§II, footnote 2] The remark that 'as a Hamiltonian itself, [S_chiral] describes decoupled wires along the z-axis of massless Majorana fermions' is illuminating and could be stated in the main text rather than in passing; it is the cleanest way to see why the spectrum is continuous.
  4. [Appendix B, Eq. (B3)] S_chiral(k) is written as a sum of four terms but no derivation or motivation is given. A brief comment on how this expression was found (e.g., a tensor product of three Su-Schrieffer-Heeger-like factors) would aid reproducibility.
  5. [Appendix C, Eq. (C2)] The bump function expression appears to have a typo: e^{|k|^2/(|k|^2 - w^2)} with |k|<w gives |k|^2-w^2<0, so the exponent is negative and the function decays — but as |k|→w^-, the exponent → -∞ and B→0, and at k=0 the exponent is 0 so B=1. Please confirm the sign convention in the exponent and clarify the normalization.
  6. [Discussion (§V)] The toric-code-as-U(1)-generator example at the end of §V is intriguing but cryptic. A reference or a one-sentence elaboration on why the toric-code Hamiltonian has integer spectrum (mod normalization) and how its ground-state degeneracy 'matches' a bosonic U(1)⋊T parity anomaly would help non-expert readers.
  7. [Throughout] The term 'ultralocal' is defined in a footnote as 'finite range in real space' on page 1; consider promoting this definition to the main text given how often it is used and contrasted with 'almost-local' (faster-than-polynomial decay) introduced in Appendix C.
  8. [Refs] Refs [4] and [8] are the same paper (Armitage-Mele-Vishwanath RMP 2018); please consolidate.

Simulated Author's Rebuttal

4 responses · 2 unresolved

We thank the referee for a careful and constructive report. The four major comments all target genuine weaknesses in exposition rather than the substance of the constructions, and we are happy to address each in revision. In summary: (1) we will replace the one-sentence protection argument in §IV with an explicit case analysis of allowed η^z σ^j terms against the linearized kinetic structures at ±K, plus a small numerical check; (2) we will temper the abstract language from \"gives rise to\" the SU(2) Witten anomaly to a more careful statement, and add a concrete consistency check showing that two copies of the doublet admit a symmetric gap, matching the mod-2 character; (3) we will reformat §III as a formal Proposition with hypotheses (H1)–(H4) and explain that (H4) — quantized charge — is precisely what fails for the §II generator, removing any apparent tension; and (4) we will rewrite §II to make explicit that Eq. (7) generates an R action that is the diagonal of two non-commuting U(1)s S_0, S_1 of §IV, which is indeed the conceptual point of the paper. We list two items as standing — a UV finite-volume invariant for Witten's anomaly built from the Onsager generators, and a proof that time reversal genuinely forces almost-locality — which we flag as open.

read point-by-point responses

  1. Referee: §IV, Eq. (14): The protection argument against translation-breaking perturbations is asserted in one sentence; please supply an explicit case analysis or numerical demonstration that no symmetric mass m_j(k) η^z σ^j gaps both valleys.

    Authors: We agree this load-bearing step deserves an explicit treatment and will expand it in the revision. The argument we have in mind is the following case analysis. The generators in (14) commute with the four matrices η^z σ^x, η^z σ^y, η^z σ^z, and η^z 1 (modulo k-dependent dressings); these are the only inter-valley bilinears compatible with both U(1)_{Q_0} and U(1)_{Q_1} for generic K. Linearizing around k=±K (Eq. 12), the kinetic structure at the two valleys is k_x σ_x + k_y σ_z ± sin K k_z σ_y. A term of the form m_j η^z σ^j either (i) anticommutes with the kinetic part at one valley but commutes at the other, in which case it merely shifts the location of one node, or (ii) commutes with at least one σ-structure at each valley, in which case it cannot open a gap but only deforms the dispersion. The η^z 1 term shifts the chemical potential of opposite-charge valleys oppositely. We will (a) tabulate this case analysis explicitly in §IV, and (b) supplement it with a numerical sweep over the symmetric perturbations on a finite lattice, confirming that a gap only opens once the perturbation is large enough to drive the nodes to K=0 or π, where the two generators align and the SU(2) collapses to U(1) — exactly as claimed. revision: yes

  2. Referee: §IV/§V: The identification of the UV Onsager algebra with the IR SU(2) Witten anomaly is asserted by analogy but not derived. Please demonstrate this concretely (e.g. a 2-doublet version that can be symmetrically gapped, or a finite-volume invariant detecting Witten's anomaly).

    Authors: This is a fair criticism, and we will soften the abstract wording from "gives rise to" to "is consistent with and emanates" the SU(2) anomaly, while adding a more concrete diagnostic in §V. Concretely: (i) two copies of our doublet model can be gapped while preserving both U(1)_{Q_0} and U(1)_{Q_1} (and translations) by pair-condensing inter-doublet masses that respect both quantized generators; we will exhibit such a mass term explicitly. This matches the mod-2 character of Witten's anomaly. (ii) For a single doublet, we can argue at the level of the IR: τ^z and cos K τ^z + sin K τ^x become two su(2) generators at angle K, so on the two-Weyl Hilbert space the anomaly is the standard Witten Z_2 of an SU(2) doublet of Weyl fermions. We will not, in this version, compute a fully UV finite-volume Z_2 invariant from the Onsager generators — that is an interesting open question we will flag in §V. We thank the referee for prompting the 2-copy check, which we agree is the cleanest available consistency test. revision: yes

  3. Referee: §III: State the no-go theorem as a formal proposition with hypotheses (translation invariance, quadratic generator, exponential locality). Clarify the role of the spectral gap above the zero eigenspace and reconcile with the gapless S_chiral of §II.

    Authors: We will reformat §III as a formal Proposition with explicit hypotheses: (H1) lattice translation invariance, (H2) S_chiral is a single-particle (quadratic) Hermitian operator, (H3) S_chiral is exponentially localized in real space, and (H4) the second-quantized generator Q̂_chiral has integer (quantized) spectrum, equivalent to S_chiral having integer single-particle spectrum. Under (H1)–(H4), S_chiral automatically has a uniform gap of 1 between any zero eigenspace and the next band, which is the input Appendix E uses to conclude exponential locality of the spectral projector P(k); we will state this gap explicitly as a derived consequence of (H4) rather than an extra assumption. The §II model evades the proposition precisely because its S_chiral has continuous spectrum on [0,1] — i.e. (H4) fails — so there is no uniform gap above zero and no contradiction. We will add a sentence at the start of §III and at the end of Appendix E making this logical flow explicit. revision: yes

  4. Referee: §II vs §III: S_chiral(k) in Eq. (7) has continuous spectrum [0,1] and so is an R action rather than a U(1) generator. Please state more sharply the relation to the quantized but non-commuting Q_0, Q_1 in Eq. (10).

    Authors: We agree this point deserves to be made more precisely and is in fact one of the conceptual punchlines. We will revise the language in §II and the Discussion as follows. The §II generator (Eq. 7) is Hermitian but its single-particle spectrum is the continuous interval [0,1] (a bundle of critical Majorana wires along ẑ); exponentiating it gives a non-compact R action, not a U(1). It is a "chiral symmetry" only in the sense that it commutes with the Hamiltonian and acts as the standard τ^z U(1) charge at the Weyl node — i.e. it emanates the IR axial U(1), but is not itself a UV U(1). The §IV decomposition S_chiral = (1/2)(S_0 + S_1) (cf. Eqs. 7, 10) shows that this R action is precisely the diagonal of two non-commuting compact U(1)s, S_0 and S_1, whose algebra is the Onsager algebra. We will rewrite the relevant paragraphs in §II and the opening of §IV to state this relationship explicitly, and adjust the abstract / introduction language to avoid implying that §II carries a UV U(1). revision: yes

standing simulated objections not resolved
  • A fully UV (finite-volume, lattice-level) Z_2 invariant computed directly from the Onsager generators Q_0, Q_1 detecting Witten's SU(2) anomaly is not provided in this work and we do not claim to have one; we flag it as an open problem in the revised Discussion.
  • We do not have a proof that time-reversal symmetric realizations (Appendices C, D) genuinely require relaxing ultralocality to almost-locality; we state this openly in §V and leave it as an open question.

Circularity Check

0 steps flagged

No significant circularity: constructions are explicit Hamiltonians with explicitly checked commutators; no fitted-parameter-as-prediction or load-bearing self-citation.

full rationale

The paper's derivation chain is constructive and self-contained. The authors write down explicit Bloch Hamiltonians (Eqs. 6, 11, B2, C3, D2) and explicit symmetry generators (Eqs. 7, 10, B3, C4, D3), and the load-bearing claim that [S_chiral(k), h^BdG(k)] = 0 is a finite matrix computation the reader can verify directly — no fitted parameter is renamed as a prediction, and no self-citation supplies a "uniqueness" input. The no-go theorem in Section III is proved from stated assumptions (translation invariance, exponential locality, integer spectrum) using a standard Riesz-projector argument (Appendix E); it does not lean on prior work by the authors. The relation to the Fidkowski–Xu theorem [22] and to Ginsparg–Wilson [13,14] is cited as context, not as a load-bearing premise. The Onsager-algebra identification (Eq. 10) is a direct algebraic observation about the explicit generators τ^z and cos(k_z)τ^z + sin(k_z)τ^x, not an imported result. The skeptic's concerns (the "shifts not gaps" argument after Eq. 14; the Onsager↔Witten Z₂ anomaly matching being structural rather than derived) are correctness/completeness gaps — the argument is sketched rather than exhausted — but they are not circularity: the claims are not defined to be true, nor are they fits to the data they predict. They could in principle be falsified by exhibiting a symmetric mass term, which is an external check. Self-citations to [11,12,33] (Gioia et al.) supply background on Weyl semimetal anomalies but are not invoked as uniqueness theorems forcing the construction. Score 1 reflects only routine background self-citation.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Model omitted the axiom ledger; defaulted for pipeline continuity.

pith-pipeline@v0.9.0 · 9711 in / 5516 out tokens · 85677 ms · 2026-05-06T21:26:30.241349+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Lattice chiral symmetry from bosons in 3+1d

    hep-th 2026-04 unverdicted novelty 7.0

    A bosonic lattice model realizes exact chiral symmetry and its anomaly in 3+1d, with the continuum limit a compact boson theory with axion-like coupling.

  2. Taste-splitting mass and edge modes in $3+1$ D staggered fermions

    hep-lat 2026-04 unverdicted novelty 7.0

    A kink in a one-link mass term for 3+1D staggered fermions creates a 2+1D domain wall with two-flavor massless Dirac fermions protected by SU(2) and parity, realizing the parity anomaly from the UV lattice Hamiltonian.